©Maki Naro

One of the difficult tasks scientists in the field of machine learning is teaching machines how to perceive and understand the visual world. But Devi Parikh, assistant professor in the Bradley Department of Electrical and Computer Engineering at Virginia Tech, has an elegant, crowd-sourced solution. She wants to use cartoons to literally illustrate tasks that would normally be difficult to describe, and Google has given her $92,000 to help her do it. She’ll be working alongside Google engineers and thousands of online citizen scientists by having them manipulate clipart images to mimic photographs, thus helping machines learn how to interpret objects and people in space. This is actually the second time she has won a Faculty Research Award. The first time was in 2012 for her research on relative attributes-based feedback for image searches.

While I was drawing today’s comic, I kept thinking, “Why do cartoon characters only have four fingers?” The Simpsons, Mickey Mouse, Calvin, Hobbes. They all only have four fingers. Why is that?

Turns out there are many answers, some say it comes from having less fingers to animate. I was always told it was to avoid big cartoony fingers forming giant cartoony hands. But much of it these days likely boils down to plain tradition. When I was a kid, all my characters had four fingers because that’s how Bill Watterson drew them, and no doubt he probably got it from somewhere as well.

For more on that topic, check out this Numberphile video by Brady Haran (you may know him from Periodic Table of Videos) where author Simon Singh explains the nature of Pi, and why the Simpsons don’t have a base-8 number system.